Your Large Vision-Language Model Only Needs A Few Attention Heads For Visual Grounding.

Seil Kang, Jinyeong Kim, Junhyeok Kim, Seong Jae Hwang

Browse the full CVPR paper archive.